What a backtest proves
A backtest turns a vague idea into a sequence of historical events. Instead of saying “buy when funding looks stressed,” the rule has to say which venue, which contract, what funding threshold, what price filter, and what happens after the trigger.
Good backtests answer practical questions:
- How many times did the rule trigger?
- Were results concentrated in one regime or spread across many periods?
- Did the rule still work after fees and realistic slippage?
- Did the rule fail during specific events, such as liquidations, exchange outages, or macro releases?
Common backtest mistakes
The biggest mistake is lookahead bias: using information that was not available at the time of the trade. A macro transcript, news article, or prediction-market move should only affect the rule after the market could have seen it.
Other common mistakes:
- Testing too many variations and keeping only the winner.
- Ignoring failed triggers because they do not fit the story.
- Using spot data to test a perpetual futures rule without funding or liquidation context.
- Treating a tiny number of triggers as statistically meaningful.
How traders should use backtests
Use a backtest to decide whether a rule is worth monitoring, not whether it deserves blind execution. The strongest workflow is: define the rule, replay it, inspect each trigger, start with alerts, then decide whether any execution should be preview-confirmed or automated.
In onchain markets, this is especially useful because venue data is often rich, timestamped, and auditable. The same discipline also applies as more traditional market workflows move onto shared rails.